Hugging Face has taken a significant step toward making robotics accessible to everyone. Its new project, SmolVLA, is a compact and fully open AI model designed for controlling robots. It can work effectively on a regular laptop, without the need for supercomputers.
The model has 450 million parameters and was trained on a collection of open datasets assembled by the LeRobot community, helping it achieve high flexibility and adaptability. One of SmolVLA’s key features is support for asynchronous processing. This means that a robot controlled by such a model can perceive information and react to changes faster. This is especially important for home and educational robotics, where accuracy matters just as much as speed.
Hugging Face has long been developing an ecosystem for accessible robotics, making open solutions for robot control increasingly popular. SmolVLA may be the turning point that truly brings robotics to the mass audience—making it understandable, accessible, and inspiring.
Source: TechCrunch
